Past Event: Oden Institute Seminar

Data-driven modeling for parametric systems with applications to uncertainty quantification for large-scale environmental problems

Dr. Hannah Lu, MIT

Abstract

Data-driven modeling is often used to describe the behavior of complex systems, whose simulation with a full model is too expensive, or to extract salient features from the full model's output. In recent work, we proposed a new physics-aware data-driven framework to construct reduced-order models of parametric complex systems. It combines the advantages of the data-driven modeling tool Dynamic Mode Decomposition (DMD) and reduced-order bases (ROBs) used for manifold interpolations in the parameter space. Attractive features of this framework include (1) the ability to handle quantities of interest (QoIs) directly, without having to access the high-fidelity models for the underlying high-dimensional state variables, as required in conventional projection-based methods such as Proper Orthogonal Decomposition (POD); (2) the improved accuracy relative to that of the conventional pure-data driven techniques such as Gaussian Processes; and (3) the low training data requirements relative to data-intensive nonlinear machine learning methods such as deep neural networks. Then I will present an application example where recent developments in data-driven modeling for parametric systems have been adapted to accelerating probabilistic assessments in large-scale environmental problems. The objective is to develop a convenient computing toolbox to provide more accurate scientific information for better environmental management and guide decision-making for stakeholders.

Biography

Hannah Lu is a postdoc associate at MIT, affiliated with the Department of Aeronautics and Astronautics, Department of Civil Environmental Engineering, Earth Resources Laboratory and Laboratory for Information and Decision Systems. She obtained her Ph.D. from Energy Science and Engineering at Stanford Doerr School of Sustainability. Her research interests lie in the field of scientific computing, reduced order modeling, uncertainty quantification and machine learning in applications of environmental fluid mechanics. She received EDGE Doctoral Fellowship, Frank G. Miller Fellowship Award and Henry J. Ramey, Jr. Fellowship Award from Stanford University, Student Travel Award from SIAM Conference on UQ, NSF Fellowship from MMLDT-CSET Conference, Travel Grant from NSF-funded HydroML Symposium, and a first-place USNCCM17 Best Presentation Award in postdoc category.